Transformers have demonstrated great potential in computer vision tasks. To avoid dense computations of self-attentions in high-resolution visual data, some recent Transformer models adopt a hierarchical design, where self-attentions are only computed within local windows. This design significantly improves the efficiency but lacks global feature reasoning in early stages. In this work, we design a multi-path structure of the Transformer, which enables local-to-global reasoning at multiple granularities in each stage. The proposed framework is computationally efficient and highly effective. With a marginal increasement in computational overhead, our model achieves notable improvements in both image classification and semantic segmentation. Code is available at https://github.com/ljpadam/LG-Transformer
@article{arxiv.2107.04735,
title = {Local-to-Global Self-Attention in Vision Transformers},
author = {Jinpeng Li and Yichao Yan and Shengcai Liao and Xiaokang Yang and Ling Shao},
journal= {arXiv preprint arXiv:2107.04735},
year = {2021}
}